diff options
Diffstat (limited to 'net/tipc/socket.c')
| -rw-r--r-- | net/tipc/socket.c | 2077 |
1 files changed, 1211 insertions, 866 deletions
diff --git a/net/tipc/socket.c b/net/tipc/socket.c index 22909036b9b..ef0475568f9 100644 --- a/net/tipc/socket.c +++ b/net/tipc/socket.c @@ -1,8 +1,8 @@ /* * net/tipc/socket.c: TIPC socket API * - * Copyright (c) 2001-2007, Ericsson AB - * Copyright (c) 2004-2007, Wind River Systems + * Copyright (c) 2001-2007, 2012-2014, Ericsson AB + * Copyright (c) 2004-2008, 2010-2013, Wind River Systems * All rights reserved. * * Redistribution and use in source and binary forms, with or without @@ -34,200 +34,255 @@ * POSSIBILITY OF SUCH DAMAGE. */ -#include <linux/module.h> -#include <linux/types.h> -#include <linux/net.h> -#include <linux/socket.h> -#include <linux/errno.h> -#include <linux/mm.h> -#include <linux/slab.h> -#include <linux/poll.h> -#include <linux/fcntl.h> -#include <asm/semaphore.h> -#include <asm/string.h> -#include <asm/atomic.h> -#include <net/sock.h> - -#include <linux/tipc.h> -#include <linux/tipc_config.h> -#include <net/tipc/tipc_msg.h> -#include <net/tipc/tipc_port.h> - #include "core.h" +#include "port.h" +#include "node.h" + +#include <linux/export.h> #define SS_LISTENING -1 /* socket is listening */ #define SS_READY -2 /* socket is connectionless */ -#define OVERLOAD_LIMIT_BASE 5000 - -struct tipc_sock { - struct sock sk; - struct tipc_port *p; - struct semaphore sem; -}; - -#define tipc_sk(sk) ((struct tipc_sock*)sk) +#define CONN_TIMEOUT_DEFAULT 8000 /* default connect timeout = 8s */ -static u32 dispatch(struct tipc_port *tport, struct sk_buff *buf); -static void wakeupdispatch(struct tipc_port *tport); +static int tipc_backlog_rcv(struct sock *sk, struct sk_buff *skb); +static void tipc_data_ready(struct sock *sk); +static void tipc_write_space(struct sock *sk); +static int tipc_release(struct socket *sock); +static int tipc_accept(struct socket *sock, struct socket *new_sock, int flags); static const struct proto_ops packet_ops; static const struct proto_ops stream_ops; static const struct proto_ops msg_ops; static struct proto tipc_proto; - -static int sockets_enabled = 0; - -static atomic_t tipc_queue_size = ATOMIC_INIT(0); - +static struct proto tipc_proto_kern; /* - * sock_lock(): Lock a port/socket pair. lock_sock() can - * not be used here, since the same lock must protect ports - * with non-socket interfaces. - * See net.c for description of locking policy. + * Revised TIPC socket locking policy: + * + * Most socket operations take the standard socket lock when they start + * and hold it until they finish (or until they need to sleep). Acquiring + * this lock grants the owner exclusive access to the fields of the socket + * data structures, with the exception of the backlog queue. A few socket + * operations can be done without taking the socket lock because they only + * read socket information that never changes during the life of the socket. + * + * Socket operations may acquire the lock for the associated TIPC port if they + * need to perform an operation on the port. If any routine needs to acquire + * both the socket lock and the port lock it must take the socket lock first + * to avoid the risk of deadlock. + * + * The dispatcher handling incoming messages cannot grab the socket lock in + * the standard fashion, since invoked it runs at the BH level and cannot block. + * Instead, it checks to see if the socket lock is currently owned by someone, + * and either handles the message itself or adds it to the socket's backlog + * queue; in the latter case the queued message is processed once the process + * owning the socket lock releases it. + * + * NOTE: Releasing the socket lock while an operation is sleeping overcomes + * the problem of a blocked socket operation preventing any other operations + * from occurring. However, applications must be careful if they have + * multiple threads trying to send (or receive) on the same socket, as these + * operations might interfere with each other. For example, doing a connect + * and a receive at the same time might allow the receive to consume the + * ACK message meant for the connect. While additional work could be done + * to try and overcome this, it doesn't seem to be worthwhile at the present. + * + * NOTE: Releasing the socket lock while an operation is sleeping also ensures + * that another operation that must be performed in a non-blocking manner is + * not delayed for very long because the lock has already been taken. + * + * NOTE: This code assumes that certain fields of a port/socket pair are + * constant over its lifetime; such fields can be examined without taking + * the socket lock and/or port lock, and do not need to be re-read even + * after resuming processing after waiting. These fields include: + * - socket type + * - pointer to socket sk structure (aka tipc_sock structure) + * - pointer to port structure + * - port reference */ -static void sock_lock(struct tipc_sock* tsock) -{ - spin_lock_bh(tsock->p->lock); -} -/* - * sock_unlock(): Unlock a port/socket pair - */ -static void sock_unlock(struct tipc_sock* tsock) -{ - spin_unlock_bh(tsock->p->lock); -} +#include "socket.h" /** - * pollmask - determine the current set of poll() events for a socket - * @sock: socket structure - * - * TIPC sets the returned events as follows: - * a) POLLRDNORM and POLLIN are set if the socket's receive queue is non-empty - * or if a connection-oriented socket is does not have an active connection - * (i.e. a read operation will not block). - * b) POLLOUT is set except when a socket's connection has been terminated - * (i.e. a write operation will not block). - * c) POLLHUP is set when a socket's connection has been terminated. - * - * IMPORTANT: The fact that a read or write operation will not block does NOT - * imply that the operation will succeed! + * advance_rx_queue - discard first buffer in socket receive queue * - * Returns pollmask value + * Caller must hold socket lock */ - -static u32 pollmask(struct socket *sock) +static void advance_rx_queue(struct sock *sk) { - u32 mask; - - if ((skb_queue_len(&sock->sk->sk_receive_queue) != 0) || - (sock->state == SS_UNCONNECTED) || - (sock->state == SS_DISCONNECTING)) - mask = (POLLRDNORM | POLLIN); - else - mask = 0; - - if (sock->state == SS_DISCONNECTING) - mask |= POLLHUP; - else - mask |= POLLOUT; - - return mask; + kfree_skb(__skb_dequeue(&sk->sk_receive_queue)); } - /** - * advance_queue - discard first buffer in queue - * @tsock: TIPC socket + * reject_rx_queue - reject all buffers in socket receive queue + * + * Caller must hold socket lock */ - -static void advance_queue(struct tipc_sock *tsock) +static void reject_rx_queue(struct sock *sk) { - sock_lock(tsock); - buf_discard(skb_dequeue(&tsock->sk.sk_receive_queue)); - sock_unlock(tsock); - atomic_dec(&tipc_queue_size); + struct sk_buff *buf; + + while ((buf = __skb_dequeue(&sk->sk_receive_queue))) + tipc_reject_msg(buf, TIPC_ERR_NO_PORT); } /** - * tipc_create - create a TIPC socket + * tipc_sk_create - create a TIPC socket + * @net: network namespace (must be default network) * @sock: pre-allocated socket structure * @protocol: protocol indicator (must be 0) + * @kern: caused by kernel or by userspace? * - * This routine creates and attaches a 'struct sock' to the 'struct socket', - * then create and attaches a TIPC port to the 'struct sock' part. + * This routine creates additional data structures used by the TIPC socket, + * initializes them, and links them together. * * Returns 0 on success, errno otherwise */ -static int tipc_create(struct net *net, struct socket *sock, int protocol) +static int tipc_sk_create(struct net *net, struct socket *sock, + int protocol, int kern) { - struct tipc_sock *tsock; - struct tipc_port *port; + const struct proto_ops *ops; + socket_state state; struct sock *sk; + struct tipc_sock *tsk; + struct tipc_port *port; u32 ref; - if (net != &init_net) - return -EAFNOSUPPORT; - + /* Validate arguments */ if (unlikely(protocol != 0)) return -EPROTONOSUPPORT; - ref = tipc_createport_raw(NULL, &dispatch, &wakeupdispatch, TIPC_LOW_IMPORTANCE); - if (unlikely(!ref)) - return -ENOMEM; - - sock->state = SS_UNCONNECTED; - switch (sock->type) { case SOCK_STREAM: - sock->ops = &stream_ops; + ops = &stream_ops; + state = SS_UNCONNECTED; break; case SOCK_SEQPACKET: - sock->ops = &packet_ops; + ops = &packet_ops; + state = SS_UNCONNECTED; break; case SOCK_DGRAM: - tipc_set_portunreliable(ref, 1); - /* fall through */ case SOCK_RDM: - tipc_set_portunreturnable(ref, 1); - sock->ops = &msg_ops; - sock->state = SS_READY; + ops = &msg_ops; + state = SS_READY; break; default: - tipc_deleteport(ref); return -EPROTOTYPE; } - sk = sk_alloc(net, AF_TIPC, GFP_KERNEL, &tipc_proto); - if (!sk) { - tipc_deleteport(ref); + /* Allocate socket's protocol area */ + if (!kern) + sk = sk_alloc(net, AF_TIPC, GFP_KERNEL, &tipc_proto); + else + sk = sk_alloc(net, AF_TIPC, GFP_KERNEL, &tipc_proto_kern); + + if (sk == NULL) + return -ENOMEM; + + tsk = tipc_sk(sk); + port = &tsk->port; + + ref = tipc_port_init(port, TIPC_LOW_IMPORTANCE); + if (!ref) { + pr_warn("Socket registration failed, ref. table exhausted\n"); + sk_free(sk); return -ENOMEM; } + /* Finish initializing socket data structures */ + sock->ops = ops; + sock->state = state; + sock_init_data(sock, sk); - init_waitqueue_head(sk->sk_sleep); - sk->sk_rcvtimeo = 8 * HZ; /* default connect timeout = 8s */ + sk->sk_backlog_rcv = tipc_backlog_rcv; + sk->sk_rcvbuf = sysctl_tipc_rmem[1]; + sk->sk_data_ready = tipc_data_ready; + sk->sk_write_space = tipc_write_space; + tsk->conn_timeout = CONN_TIMEOUT_DEFAULT; + atomic_set(&tsk->dupl_rcvcnt, 0); + tipc_port_unlock(port); + + if (sock->state == SS_READY) { + tipc_port_set_unreturnable(port, true); + if (sock->type == SOCK_DGRAM) + tipc_port_set_unreliable(port, true); + } + return 0; +} - tsock = tipc_sk(sk); - port = tipc_get_port(ref); +/** + * tipc_sock_create_local - create TIPC socket from inside TIPC module + * @type: socket type - SOCK_RDM or SOCK_SEQPACKET + * + * We cannot use sock_creat_kern here because it bumps module user count. + * Since socket owner and creator is the same module we must make sure + * that module count remains zero for module local sockets, otherwise + * we cannot do rmmod. + * + * Returns 0 on success, errno otherwise + */ +int tipc_sock_create_local(int type, struct socket **res) +{ + int rc; - tsock->p = port; - port->usr_handle = tsock; + rc = sock_create_lite(AF_TIPC, type, 0, res); + if (rc < 0) { + pr_err("Failed to create kernel socket\n"); + return rc; + } + tipc_sk_create(&init_net, *res, 0, 1); - init_MUTEX(&tsock->sem); + return 0; +} - dbg("sock_create: %x\n",tsock); +/** + * tipc_sock_release_local - release socket created by tipc_sock_create_local + * @sock: the socket to be released. + * + * Module reference count is not incremented when such sockets are created, + * so we must keep it from being decremented when they are released. + */ +void tipc_sock_release_local(struct socket *sock) +{ + tipc_release(sock); + sock->ops = NULL; + sock_release(sock); +} - atomic_inc(&tipc_user_count); +/** + * tipc_sock_accept_local - accept a connection on a socket created + * with tipc_sock_create_local. Use this function to avoid that + * module reference count is inadvertently incremented. + * + * @sock: the accepting socket + * @newsock: reference to the new socket to be created + * @flags: socket flags + */ - return 0; +int tipc_sock_accept_local(struct socket *sock, struct socket **newsock, + int flags) +{ + struct sock *sk = sock->sk; + int ret; + + ret = sock_create_lite(sk->sk_family, sk->sk_type, + sk->sk_protocol, newsock); + if (ret < 0) + return ret; + + ret = tipc_accept(sock, *newsock, flags); + if (ret < 0) { + sock_release(*newsock); + return ret; + } + (*newsock)->ops = sock->ops; + return ret; } /** - * release - destroy a TIPC socket + * tipc_release - destroy a TIPC socket * @sock: socket to destroy * * This routine cleans up any messages that are still queued on the socket. @@ -242,62 +297,64 @@ static int tipc_create(struct net *net, struct socket *sock, int protocol) * * Returns 0 on success, errno otherwise */ - -static int release(struct socket *sock) +static int tipc_release(struct socket *sock) { - struct tipc_sock *tsock = tipc_sk(sock->sk); struct sock *sk = sock->sk; - int res = TIPC_OK; + struct tipc_sock *tsk; + struct tipc_port *port; struct sk_buff *buf; - dbg("sock_delete: %x\n",tsock); - if (!tsock) - return 0; - down(&tsock->sem); - if (!sock->sk) { - up(&tsock->sem); + /* + * Exit if socket isn't fully initialized (occurs when a failed accept() + * releases a pre-allocated child socket that was never used) + */ + if (sk == NULL) return 0; - } - /* Reject unreceived messages, unless no longer connected */ + tsk = tipc_sk(sk); + port = &tsk->port; + lock_sock(sk); + /* + * Reject all unreceived messages, except on an active connection + * (which disconnects locally & sends a 'FIN+' to peer) + */ while (sock->state != SS_DISCONNECTING) { - sock_lock(tsock); - buf = skb_dequeue(&sk->sk_receive_queue); - if (!buf) - tsock->p->usr_handle = NULL; - sock_unlock(tsock); - if (!buf) + buf = __skb_dequeue(&sk->sk_receive_queue); + if (buf == NULL) break; - if (TIPC_SKB_CB(buf)->handle != msg_data(buf_msg(buf))) - buf_discard(buf); - else + if (TIPC_SKB_CB(buf)->handle != NULL) + kfree_skb(buf); + else { + if ((sock->state == SS_CONNECTING) || + (sock->state == SS_CONNECTED)) { + sock->state = SS_DISCONNECTING; + tipc_port_disconnect(port->ref); + } tipc_reject_msg(buf, TIPC_ERR_NO_PORT); - atomic_dec(&tipc_queue_size); + } } - /* Delete TIPC port */ - - res = tipc_deleteport(tsock->p->ref); - sock->sk = NULL; + /* Destroy TIPC port; also disconnects an active connection and + * sends a 'FIN-' to peer. + */ + tipc_port_destroy(port); - /* Discard any remaining messages */ + /* Discard any remaining (connection-based) messages in receive queue */ + __skb_queue_purge(&sk->sk_receive_queue); - while ((buf = skb_dequeue(&sk->sk_receive_queue))) { - buf_discard(buf); - atomic_dec(&tipc_queue_size); - } - - up(&tsock->sem); + /* Reject any messages that accumulated in backlog queue */ + sock->state = SS_DISCONNECTING; + release_sock(sk); sock_put(sk); + sock->sk = NULL; - atomic_dec(&tipc_user_count); - return res; + return 0; } /** - * bind - associate or disassocate TIPC name(s) with a socket + * tipc_bind - associate or disassocate TIPC name(s) with a socket * @sock: socket structure * @uaddr: socket address describing name(s) and desired operation * @uaddr_len: size of socket address data structure @@ -307,19 +364,21 @@ static int release(struct socket *sock) * (i.e. a socket address length of 0) unbinds all names from the socket. * * Returns 0 on success, errno otherwise + * + * NOTE: This routine doesn't need to take the socket lock since it doesn't + * access any non-constant socket information. */ - -static int bind(struct socket *sock, struct sockaddr *uaddr, int uaddr_len) +static int tipc_bind(struct socket *sock, struct sockaddr *uaddr, + int uaddr_len) { - struct tipc_sock *tsock = tipc_sk(sock->sk); + struct sock *sk = sock->sk; struct sockaddr_tipc *addr = (struct sockaddr_tipc *)uaddr; - int res; - - if (down_interruptible(&tsock->sem)) - return -ERESTARTSYS; + struct tipc_sock *tsk = tipc_sk(sk); + int res = -EINVAL; + lock_sock(sk); if (unlikely(!uaddr_len)) { - res = tipc_withdraw(tsock->p->ref, 0, NULL); + res = tipc_withdraw(&tsk->port, 0, NULL); goto exit; } @@ -327,11 +386,11 @@ static int bind(struct socket *sock, struct sockaddr *uaddr, int uaddr_len) res = -EINVAL; goto exit; } - if (addr->family != AF_TIPC) { res = -EAFNOSUPPORT; goto exit; } + if (addr->addrtype == TIPC_ADDR_NAME) addr->addr.nameseq.upper = addr->addr.nameseq.lower; else if (addr->addrtype != TIPC_ADDR_NAMESEQ) { @@ -339,66 +398,131 @@ static int bind(struct socket *sock, struct sockaddr *uaddr, int uaddr_len) goto exit; } - if (addr->scope > 0) - res = tipc_publish(tsock->p->ref, addr->scope, - &addr->addr.nameseq); - else - res = tipc_withdraw(tsock->p->ref, -addr->scope, - &addr->addr.nameseq); + if ((addr->addr.nameseq.type < TIPC_RESERVED_TYPES) && + (addr->addr.nameseq.type != TIPC_TOP_SRV) && + (addr->addr.nameseq.type != TIPC_CFG_SRV)) { + res = -EACCES; + goto exit; + } + + res = (addr->scope > 0) ? + tipc_publish(&tsk->port, addr->scope, &addr->addr.nameseq) : + tipc_withdraw(&tsk->port, -addr->scope, &addr->addr.nameseq); exit: - up(&tsock->sem); + release_sock(sk); return res; } /** - * get_name - get port ID of socket or peer socket + * tipc_getname - get port ID of socket or peer socket * @sock: socket structure * @uaddr: area for returned socket address * @uaddr_len: area for returned length of socket address - * @peer: 0 to obtain socket name, 1 to obtain peer socket name + * @peer: 0 = own ID, 1 = current peer ID, 2 = current/former peer ID * * Returns 0 on success, errno otherwise + * + * NOTE: This routine doesn't need to take the socket lock since it only + * accesses socket information that is unchanging (or which changes in + * a completely predictable manner). */ - -static int get_name(struct socket *sock, struct sockaddr *uaddr, - int *uaddr_len, int peer) +static int tipc_getname(struct socket *sock, struct sockaddr *uaddr, + int *uaddr_len, int peer) { - struct tipc_sock *tsock = tipc_sk(sock->sk); struct sockaddr_tipc *addr = (struct sockaddr_tipc *)uaddr; - u32 res; + struct tipc_sock *tsk = tipc_sk(sock->sk); - if (down_interruptible(&tsock->sem)) - return -ERESTARTSYS; + memset(addr, 0, sizeof(*addr)); + if (peer) { + if ((sock->state != SS_CONNECTED) && + ((peer != 2) || (sock->state != SS_DISCONNECTING))) + return -ENOTCONN; + addr->addr.id.ref = tipc_port_peerport(&tsk->port); + addr->addr.id.node = tipc_port_peernode(&tsk->port); + } else { + addr->addr.id.ref = tsk->port.ref; + addr->addr.id.node = tipc_own_addr; + } *uaddr_len = sizeof(*addr); addr->addrtype = TIPC_ADDR_ID; addr->family = AF_TIPC; addr->scope = 0; - if (peer) - res = tipc_peer(tsock->p->ref, &addr->addr.id); - else - res = tipc_ownidentity(tsock->p->ref, &addr->addr.id); addr->addr.name.domain = 0; - up(&tsock->sem); - return res; + return 0; } /** - * poll - read and possibly block on pollmask + * tipc_poll - read and possibly block on pollmask * @file: file structure associated with the socket * @sock: socket for which to calculate the poll bits * @wait: ??? * - * Returns the pollmask + * Returns pollmask value + * + * COMMENTARY: + * It appears that the usual socket locking mechanisms are not useful here + * since the pollmask info is potentially out-of-date the moment this routine + * exits. TCP and other protocols seem to rely on higher level poll routines + * to handle any preventable race conditions, so TIPC will do the same ... + * + * TIPC sets the returned events as follows: + * + * socket state flags set + * ------------ --------- + * unconnected no read flags + * POLLOUT if port is not congested + * + * connecting POLLIN/POLLRDNORM if ACK/NACK in rx queue + * no write flags + * + * connected POLLIN/POLLRDNORM if data in rx queue + * POLLOUT if port is not congested + * + * disconnecting POLLIN/POLLRDNORM/POLLHUP + * no write flags + * + * listening POLLIN if SYN in rx queue + * no write flags + * + * ready POLLIN/POLLRDNORM if data in rx queue + * [connectionless] POLLOUT (since port cannot be congested) + * + * IMPORTANT: The fact that a read or write operation is indicated does NOT + * imply that the operation will succeed, merely that it should be performed + * and will not block. */ - -static unsigned int poll(struct file *file, struct socket *sock, - poll_table *wait) +static unsigned int tipc_poll(struct file *file, struct socket *sock, + poll_table *wait) { - poll_wait(file, sock->sk->sk_sleep, wait); - /* NEED LOCK HERE? */ - return pollmask(sock); + struct sock *sk = sock->sk; + struct tipc_sock *tsk = tipc_sk(sk); + u32 mask = 0; + + sock_poll_wait(file, sk_sleep(sk), wait); + + switch ((int)sock->state) { + case SS_UNCONNECTED: + if (!tsk->port.congested) + mask |= POLLOUT; + break; + case SS_READY: + case SS_CONNECTED: + if (!tsk->port.congested) + mask |= POLLOUT; + /* fall thru' */ + case SS_CONNECTING: + case SS_LISTENING: + if (!skb_queue_empty(&sk->sk_receive_queue)) + mask |= (POLLIN | POLLRDNORM); + break; + case SS_DISCONNECTING: + mask = (POLLIN | POLLRDNORM | POLLHUP); + break; + } + + return mask; } /** @@ -411,7 +535,6 @@ static unsigned int poll(struct file *file, struct socket *sock, * * Returns 0 if permission is granted, otherwise errno */ - static int dest_name_check(struct sockaddr_tipc *dest, struct msghdr *m) { struct tipc_cfg_msg_hdr hdr; @@ -420,10 +543,11 @@ static int dest_name_check(struct sockaddr_tipc *dest, struct msghdr *m) return 0; if (likely(dest->addr.name.name.type == TIPC_TOP_SRV)) return 0; - if (likely(dest->addr.name.name.type != TIPC_CFG_SRV)) return -EACCES; + if (!m->msg_iovlen || (m->msg_iov[0].iov_len < sizeof(hdr))) + return -EMSGSIZE; if (copy_from_user(&hdr, m->msg_iov[0].iov_base, sizeof(hdr))) return -EFAULT; if ((ntohs(hdr.tcm_type) & 0xC000) && (!capable(CAP_NET_ADMIN))) @@ -432,9 +556,35 @@ static int dest_name_check(struct sockaddr_tipc *dest, struct msghdr *m) return 0; } +static int tipc_wait_for_sndmsg(struct socket *sock, long *timeo_p) +{ + struct sock *sk = sock->sk; + struct tipc_sock *tsk = tipc_sk(sk); + DEFINE_WAIT(wait); + int done; + + do { + int err = sock_error(sk); + if (err) + return err; + if (sock->state == SS_DISCONNECTING) + return -EPIPE; + if (!*timeo_p) + return -EAGAIN; + if (signal_pending(current)) + return sock_intr_errno(*timeo_p); + + prepare_to_wait(sk_sleep(sk), &wait, TASK_INTERRUPTIBLE); + done = sk_wait_event(sk, timeo_p, !tsk->port.congested); + finish_wait(sk_sleep(sk), &wait); + } while (!done); + return 0; +} + + /** - * send_msg - send message in connectionless manner - * @iocb: (unused) + * tipc_sendmsg - send message in connectionless manner + * @iocb: if NULL, indicates that socket lock is already held * @sock: socket structure * @m: message to send * @total_len: length of message @@ -446,14 +596,15 @@ static int dest_name_check(struct sockaddr_tipc *dest, struct msghdr *m) * * Returns the number of bytes sent on success, or errno otherwise */ - -static int send_msg(struct kiocb *iocb, struct socket *sock, - struct msghdr *m, size_t total_len) +static int tipc_sendmsg(struct kiocb *iocb, struct socket *sock, + struct msghdr *m, size_t total_len) { - struct tipc_sock *tsock = tipc_sk(sock->sk); - struct sockaddr_tipc *dest = (struct sockaddr_tipc *)m->msg_name; - struct sk_buff *buf; + struct sock *sk = sock->sk; + struct tipc_sock *tsk = tipc_sk(sk); + struct tipc_port *port = &tsk->port; + DECLARE_SOCKADDR(struct sockaddr_tipc *, dest, m->msg_name); int needs_conn; + long timeo; int res = -EINVAL; if (unlikely(!dest)) @@ -461,86 +612,112 @@ static int send_msg(struct kiocb *iocb, struct socket *sock, if (unlikely((m->msg_namelen < sizeof(*dest)) || (dest->family != AF_TIPC))) return -EINVAL; + if (total_len > TIPC_MAX_USER_MSG_SIZE) + return -EMSGSIZE; + + if (iocb) + lock_sock(sk); needs_conn = (sock->state != SS_READY); if (unlikely(needs_conn)) { - if (sock->state == SS_LISTENING) - return -EPIPE; - if (sock->state != SS_UNCONNECTED) - return -EISCONN; - if ((tsock->p->published) || - ((sock->type == SOCK_STREAM) && (total_len != 0))) - return -EOPNOTSUPP; + if (sock->state == SS_LISTENING) { + res = -EPIPE; + goto exit; + } + if (sock->state != SS_UNCONNECTED) { + res = -EISCONN; + goto exit; + } + if (tsk->port.published) { + res = -EOPNOTSUPP; + goto exit; + } if (dest->addrtype == TIPC_ADDR_NAME) { - tsock->p->conn_type = dest->addr.name.name.type; - tsock->p->conn_instance = dest->addr.name.name.instance; + tsk->port.conn_type = dest->addr.name.name.type; + tsk->port.conn_instance = dest->addr.name.name.instance; } - } - - if (down_interruptible(&tsock->sem)) - return -ERESTARTSYS; - - if (needs_conn) { /* Abort any pending connection attempts (very unlikely) */ - - while ((buf = skb_dequeue(&sock->sk->sk_receive_queue))) { - tipc_reject_msg(buf, TIPC_ERR_NO_PORT); - atomic_dec(&tipc_queue_size); - } - - sock->state = SS_CONNECTING; + reject_rx_queue(sk); } + timeo = sock_sndtimeo(sk, m->msg_flags & MSG_DONTWAIT); do { if (dest->addrtype == TIPC_ADDR_NAME) { - if ((res = dest_name_check(dest, m))) - goto exit; - res = tipc_send2name(tsock->p->ref, + res = dest_name_check(dest, m); + if (res) + break; + res = tipc_send2name(port, &dest->addr.name.name, dest->addr.name.domain, - m->msg_iovlen, - m->msg_iov); - } - else if (dest->addrtype == TIPC_ADDR_ID) { - res = tipc_send2port(tsock->p->ref, + m->msg_iov, + total_len); + } else if (dest->addrtype == TIPC_ADDR_ID) { + res = tipc_send2port(port, &dest->addr.id, - m->msg_iovlen, - m->msg_iov); - } - else if (dest->addrtype == TIPC_ADDR_MCAST) { + m->msg_iov, + total_len); + } else if (dest->addrtype == TIPC_ADDR_MCAST) { if (needs_conn) { res = -EOPNOTSUPP; - goto exit; + break; } - if ((res = dest_name_check(dest, m))) - goto exit; - res = tipc_multicast(tsock->p->ref, - &dest->addr.nameseq, - 0, - m->msg_iovlen, - m->msg_iov); + res = dest_name_check(dest, m); + if (res) + break; + res = tipc_port_mcast_xmit(port, + &dest->addr.nameseq, + m->msg_iov, + total_len); } if (likely(res != -ELINKCONG)) { -exit: - up(&tsock->sem); - return res; - } - if (m->msg_flags & MSG_DONTWAIT) { - res = -EWOULDBLOCK; - goto exit; - } - if (wait_event_interruptible(*sock->sk->sk_sleep, - !tsock->p->congested)) { - res = -ERESTARTSYS; - goto exit; + if (needs_conn && (res >= 0)) + sock->state = SS_CONNECTING; + break; } + res = tipc_wait_for_sndmsg(sock, &timeo); + if (res) + break; } while (1); + +exit: + if (iocb) + release_sock(sk); + return res; +} + +static int tipc_wait_for_sndpkt(struct socket *sock, long *timeo_p) +{ + struct sock *sk = sock->sk; + struct tipc_sock *tsk = tipc_sk(sk); + struct tipc_port *port = &tsk->port; + DEFINE_WAIT(wait); + int done; + + do { + int err = sock_error(sk); + if (err) + return err; + if (sock->state == SS_DISCONNECTING) + return -EPIPE; + else if (sock->state != SS_CONNECTED) + return -ENOTCONN; + if (!*timeo_p) + return -EAGAIN; + if (signal_pending(current)) + return sock_intr_errno(*timeo_p); + + prepare_to_wait(sk_sleep(sk), &wait, TASK_INTERRUPTIBLE); + done = sk_wait_event(sk, timeo_p, + (!port->congested || !port->connected)); + finish_wait(sk_sleep(sk), &wait); + } while (!done); + return 0; } /** - * send_packet - send a connection-oriented message - * @iocb: (unused) + * tipc_send_packet - send a connection-oriented message + * @iocb: if NULL, indicates that socket lock is already held * @sock: socket structure * @m: message to send * @total_len: length of message @@ -549,52 +726,50 @@ exit: * * Returns the number of bytes sent on success, or errno otherwise */ - -static int send_packet(struct kiocb *iocb, struct socket *sock, - struct msghdr *m, size_t total_len) +static int tipc_send_packet(struct kiocb *iocb, struct socket *sock, + struct msghdr *m, size_t total_len) { - struct tipc_sock *tsock = tipc_sk(sock->sk); - struct sockaddr_tipc *dest = (struct sockaddr_tipc *)m->msg_name; - int res; + struct sock *sk = sock->sk; + struct tipc_sock *tsk = tipc_sk(sk); + DECLARE_SOCKADDR(struct sockaddr_tipc *, dest, m->msg_name); + int res = -EINVAL; + long timeo; /* Handle implied connection establishment */ - if (unlikely(dest)) - return send_msg(iocb, sock, m, total_len); + return tipc_sendmsg(iocb, sock, m, total_len); + + if (total_len > TIPC_MAX_USER_MSG_SIZE) + return -EMSGSIZE; - if (down_interruptible(&tsock->sem)) { - return -ERESTARTSYS; + if (iocb) + lock_sock(sk); + + if (unlikely(sock->state != SS_CONNECTED)) { + if (sock->state == SS_DISCONNECTING) + res = -EPIPE; + else + res = -ENOTCONN; + goto exit; } + timeo = sock_sndtimeo(sk, m->msg_flags & MSG_DONTWAIT); do { - if (unlikely(sock->state != SS_CONNECTED)) { - if (sock->state == SS_DISCONNECTING) - res = -EPIPE; - else - res = -ENOTCONN; - goto exit; - } - - res = tipc_send(tsock->p->ref, m->msg_iovlen, m->msg_iov); - if (likely(res != -ELINKCONG)) { -exit: - up(&tsock->sem); - return res; - } - if (m->msg_flags & MSG_DONTWAIT) { - res = -EWOULDBLOCK; - goto exit; - } - if (wait_event_interruptible(*sock->sk->sk_sleep, - !tsock->p->congested)) { - res = -ERESTARTSYS; - goto exit; - } + res = tipc_send(&tsk->port, m->msg_iov, total_len); + if (likely(res != -ELINKCONG)) + break; + res = tipc_wait_for_sndpkt(sock, &timeo); + if (res) + break; } while (1); +exit: + if (iocb) + release_sock(sk); + return res; } /** - * send_stream - send stream-oriented data + * tipc_send_stream - send stream-oriented data * @iocb: (unused) * @sock: socket structure * @m: data to send @@ -605,12 +780,11 @@ exit: * Returns the number of bytes sent on success (or partial success), * or errno if no data sent */ - - -static int send_stream(struct kiocb *iocb, struct socket *sock, - struct msghdr *m, size_t total_len) +static int tipc_send_stream(struct kiocb *iocb, struct socket *sock, + struct msghdr *m, size_t total_len) { - struct tipc_port *tport; + struct sock *sk = sock->sk; + struct tipc_sock *tsk = tipc_sk(sk); struct msghdr my_msg; struct iovec my_iov; struct iovec *curr_iov; @@ -622,19 +796,26 @@ static int send_stream(struct kiocb *iocb, struct socket *sock, int bytes_sent; int res; - /* Handle special cases where there is no connection */ + lock_sock(sk); + /* Handle special cases where there is no connection */ if (unlikely(sock->state != SS_CONNECTED)) { if (sock->state == SS_UNCONNECTED) - return send_packet(iocb, sock, m, total_len); - else if (sock->state == SS_DISCONNECTING) - return -EPIPE; + res = tipc_send_packet(NULL, sock, m, total_len); else - return -ENOTCONN; + res = sock->state == SS_DISCONNECTING ? -EPIPE : -ENOTCONN; + goto exit; } - if (unlikely(m->msg_name)) - return -EISCONN; + if (unlikely(m->msg_name)) { + res = -EISCONN; + goto exit; + } + + if (total_len > (unsigned int)INT_MAX) { + res = -EMSGSIZE; + goto exit; + } /* * Send each iovec entry using one or more messages @@ -643,7 +824,6 @@ static int send_stream(struct kiocb *iocb, struct socket *sock, * (i.e. one large iovec entry), but could be improved to pass sets * of small iovec entries into send_packet(). */ - curr_iov = m->msg_iov; curr_iovlen = m->msg_iovlen; my_msg.msg_iov = &my_iov; @@ -652,25 +832,26 @@ static int send_stream(struct kiocb *iocb, struct socket *sock, my_msg.msg_name = NULL; bytes_sent = 0; - tport = tipc_sk(sock->sk)->p; - hdr_size = msg_hdr_sz(&tport->phdr); + hdr_size = msg_hdr_sz(&tsk->port.phdr); while (curr_iovlen--) { curr_start = curr_iov->iov_base; curr_left = curr_iov->iov_len; while (curr_left) { - bytes_to_send = tport->max_pkt - hdr_size; + bytes_to_send = tsk->port.max_pkt - hdr_size; if (bytes_to_send > TIPC_MAX_USER_MSG_SIZE) bytes_to_send = TIPC_MAX_USER_MSG_SIZE; if (curr_left < bytes_to_send) bytes_to_send = curr_left; my_iov.iov_base = curr_start; my_iov.iov_len = bytes_to_send; - if ((res = send_packet(iocb, sock, &my_msg, 0)) < 0) { - if (bytes_sent != 0) + res = tipc_send_packet(NULL, sock, &my_msg, + bytes_to_send); + if (res < 0) { + if (bytes_sent) res = bytes_sent; - return res; + goto exit; } curr_left -= bytes_to_send; curr_start += bytes_to_send; @@ -679,33 +860,33 @@ static int send_stream(struct kiocb *iocb, struct socket *sock, curr_iov++; } - - return bytes_sent; + res = bytes_sent; +exit: + release_sock(sk); + return res; } /** * auto_connect - complete connection setup to a remote port - * @sock: socket structure - * @tsock: TIPC-specific socket structure + * @tsk: tipc socket structure * @msg: peer's response message * * Returns 0 on success, errno otherwise */ - -static int auto_connect(struct socket *sock, struct tipc_sock *tsock, - struct tipc_msg *msg) +static int auto_connect(struct tipc_sock *tsk, struct tipc_msg *msg) { + struct tipc_port *port = &tsk->port; + struct socket *sock = tsk->sk.sk_socket; struct tipc_portid peer; - if (msg_errcode(msg)) { - sock->state = SS_DISCONNECTING; - return -ECONNREFUSED; - } - peer.ref = msg_origport(msg); peer.node = msg_orignode(msg); - tipc_connect2port(tsock->p->ref, &peer); - tipc_set_portimportance(tsock->p->ref, msg_importance(msg)); + + __tipc_port_connect(port->ref, port, &peer); + + if (msg_importance(msg) > TIPC_CRITICAL_IMPORTANCE) + return -EINVAL; + msg_set_importance(&port->phdr, (u32)msg_importance(msg)); sock->state = SS_CONNECTED; return 0; } @@ -717,18 +898,18 @@ static int auto_connect(struct socket *sock, struct tipc_sock *tsock, * * Note: Address is not captured if not requested by receiver. */ - static void set_orig_addr(struct msghdr *m, struct tipc_msg *msg) { - struct sockaddr_tipc *addr = (struct sockaddr_tipc *)m->msg_name; + DECLARE_SOCKADDR(struct sockaddr_tipc *, addr, m->msg_name); if (addr) { addr->family = AF_TIPC; addr->addrtype = TIPC_ADDR_ID; + memset(&addr->addr, 0, sizeof(addr->addr)); addr->addr.id.ref = msg_origport(msg); addr->addr.id.node = msg_orignode(msg); - addr->addr.name.domain = 0; /* could leave uninitialized */ - addr->scope = 0; /* could leave uninitialized */ + addr->addr.name.domain = 0; /* could leave uninitialized */ + addr->scope = 0; /* could leave uninitialized */ m->msg_namelen = sizeof(struct sockaddr_tipc); } } @@ -743,9 +924,8 @@ static void set_orig_addr(struct msghdr *m, struct tipc_msg *msg) * * Returns 0 if successful, otherwise errno */ - static int anc_data_recv(struct msghdr *m, struct tipc_msg *msg, - struct tipc_port *tport) + struct tipc_port *tport) { u32 anc_data[3]; u32 err; @@ -757,21 +937,22 @@ static int anc_data_recv(struct msghdr *m, struct tipc_msg *msg, return 0; /* Optionally capture errored message object(s) */ - err = msg ? msg_errcode(msg) : 0; if (unlikely(err)) { anc_data[0] = err; anc_data[1] = msg_data_sz(msg); - if ((res = put_cmsg(m, SOL_TIPC, TIPC_ERRINFO, 8, anc_data))) - return res; - if (anc_data[1] && - (res = put_cmsg(m, SOL_TIPC, TIPC_RETDATA, anc_data[1], - msg_data(msg)))) + res = put_cmsg(m, SOL_TIPC, TIPC_ERRINFO, 8, anc_data); + if (res) return res; + if (anc_data[1]) { + res = put_cmsg(m, SOL_TIPC, TIPC_RETDATA, anc_data[1], + msg_data(msg)); + if (res) + return res; + } } /* Optionally capture message destination object */ - dest_type = msg ? msg_type(msg) : TIPC_DIRECT_MSG; switch (dest_type) { case TIPC_NAMED_MSG: @@ -795,15 +976,50 @@ static int anc_data_recv(struct msghdr *m, struct tipc_msg *msg, default: has_name = 0; } - if (has_name && - (res = put_cmsg(m, SOL_TIPC, TIPC_DESTNAME, 12, anc_data))) - return res; + if (has_name) { + res = put_cmsg(m, SOL_TIPC, TIPC_DESTNAME, 12, anc_data); + if (res) + return res; + } return 0; } +static int tipc_wait_for_rcvmsg(struct socket *sock, long *timeop) +{ + struct sock *sk = sock->sk; + DEFINE_WAIT(wait); + long timeo = *timeop; + int err; + + for (;;) { + prepare_to_wait(sk_sleep(sk), &wait, TASK_INTERRUPTIBLE); + if (timeo && skb_queue_empty(&sk->sk_receive_queue)) { + if (sock->state == SS_DISCONNECTING) { + err = -ENOTCONN; + break; + } + release_sock(sk); + timeo = schedule_timeout(timeo); + lock_sock(sk); + } + err = 0; + if (!skb_queue_empty(&sk->sk_receive_queue)) + break; + err = sock_intr_errno(timeo); + if (signal_pending(current)) + break; + err = -EAGAIN; + if (!timeo) + break; + } + finish_wait(sk_sleep(sk), &wait); + *timeop = timeo; + return err; +} + /** - * recv_msg - receive packet-oriented message + * tipc_recvmsg - receive packet-oriented message * @iocb: (unused) * @m: descriptor for message info * @buf_len: total size of user buffer area @@ -814,105 +1030,68 @@ static int anc_data_recv(struct msghdr *m, struct tipc_msg *msg, * * Returns size of returned message data, errno otherwise */ - -static int recv_msg(struct kiocb *iocb, struct socket *sock, - struct msghdr *m, size_t buf_len, int flags) +static int tipc_recvmsg(struct kiocb *iocb, struct socket *sock, + struct msghdr *m, size_t buf_len, int flags) { - struct tipc_sock *tsock = tipc_sk(sock->sk); + struct sock *sk = sock->sk; + struct tipc_sock *tsk = tipc_sk(sk); + struct tipc_port *port = &tsk->port; struct sk_buff *buf; struct tipc_msg *msg; - unsigned int q_len; + long timeo; unsigned int sz; u32 err; int res; - /* Currently doesn't support receiving into multiple iovec entries */ - - if (m->msg_iovlen != 1) - return -EOPNOTSUPP; - - /* Catch invalid receive attempts */ - + /* Catch invalid receive requests */ if (unlikely(!buf_len)) return -EINVAL; - if (sock->type == SOCK_SEQPACKET) { - if (unlikely(sock->state == SS_UNCONNECTED)) - return -ENOTCONN; - if (unlikely((sock->state == SS_DISCONNECTING) && - (skb_queue_len(&sock->sk->sk_receive_queue) == 0))) - return -ENOTCONN; - } - - /* Look for a message in receive queue; wait if necessary */ - - if (unlikely(down_interruptible(&tsock->sem))) - return -ERESTARTSYS; - -restart: - if (unlikely((skb_queue_len(&sock->sk->sk_receive_queue) == 0) && - (flags & MSG_DONTWAIT))) { - res = -EWOULDBLOCK; - goto exit; - } + lock_sock(sk); - if ((res = wait_event_interruptible( - *sock->sk->sk_sleep, - ((q_len = skb_queue_len(&sock->sk->sk_receive_queue)) || - (sock->state == SS_DISCONNECTING))) )) { + if (unlikely(sock->state == SS_UNCONNECTED)) { + res = -ENOTCONN; goto exit; } - /* Catch attempt to receive on an already terminated connection */ - /* [THIS CHECK MAY OVERLAP WITH AN EARLIER CHECK] */ + timeo = sock_rcvtimeo(sk, flags & MSG_DONTWAIT); +restart: - if (!q_len) { - res = -ENOTCONN; + /* Look for a message in receive queue; wait if necessary */ + res = tipc_wait_for_rcvmsg(sock, &timeo); + if (res) goto exit; - } - - /* Get access to first message in receive queue */ - buf = skb_peek(&sock->sk->sk_receive_queue); + /* Look at first message in receive queue */ + buf = skb_peek(&sk->sk_receive_queue); msg = buf_msg(buf); sz = msg_data_sz(msg); err = msg_errcode(msg); - /* Complete connection setup for an implied connect */ - - if (unlikely(sock->state == SS_CONNECTING)) { - if ((res = auto_connect(sock, tsock, msg))) - goto exit; - } - /* Discard an empty non-errored message & try again */ - if ((!sz) && (!err)) { - advance_queue(tsock); + advance_rx_queue(sk); goto restart; } /* Capture sender's address (optional) */ - set_orig_addr(m, msg); /* Capture ancillary data (optional) */ - - if ((res = anc_data_recv(m, msg, tsock->p))) + res = anc_data_recv(m, msg, port); + if (res) goto exit; /* Capture message data (if valid) & compute return value (always) */ - if (!err) { if (unlikely(buf_len < sz)) { sz = buf_len; m->msg_flags |= MSG_TRUNC; } - if (unlikely(copy_to_user(m->msg_iov->iov_base, msg_data(msg), - sz))) { - res = -EFAULT; + res = skb_copy_datagram_iovec(buf, msg_hdr_sz(msg), + m->msg_iov, sz); + if (res) goto exit; - } res = sz; } else { if ((sock->state == SS_READY) || @@ -923,19 +1102,19 @@ restart: } /* Consume received message (optional) */ - if (likely(!(flags & MSG_PEEK))) { - if (unlikely(++tsock->p->conn_unacked >= TIPC_FLOW_CONTROL_WIN)) - tipc_acknowledge(tsock->p->ref, tsock->p->conn_unacked); - advance_queue(tsock); + if ((sock->state != SS_READY) && + (++port->conn_unacked >= TIPC_CONNACK_INTV)) + tipc_acknowledge(port->ref, port->conn_unacked); + advance_rx_queue(sk); } exit: - up(&tsock->sem); + release_sock(sk); return res; } /** - * recv_stream - receive stream-oriented data + * tipc_recv_stream - receive stream-oriented data * @iocb: (unused) * @m: descriptor for message info * @buf_len: total size of user buffer area @@ -946,109 +1125,82 @@ exit: * * Returns size of returned message data, errno otherwise */ - -static int recv_stream(struct kiocb *iocb, struct socket *sock, - struct msghdr *m, size_t buf_len, int flags) +static int tipc_recv_stream(struct kiocb *iocb, struct socket *sock, + struct msghdr *m, size_t buf_len, int flags) { - struct tipc_sock *tsock = tipc_sk(sock->sk); + struct sock *sk = sock->sk; + struct tipc_sock *tsk = tipc_sk(sk); + struct tipc_port *port = &tsk->port; struct sk_buff *buf; struct tipc_msg *msg; - unsigned int q_len; + long timeo; unsigned int sz; - int sz_to_copy; + int sz_to_copy, target, needed; int sz_copied = 0; - int needed; - char __user *crs = m->msg_iov->iov_base; - unsigned char *buf_crs; u32 err; - int res; - - /* Currently doesn't support receiving into multiple iovec entries */ - - if (m->msg_iovlen != 1) - return -EOPNOTSUPP; + int res = 0; /* Catch invalid receive attempts */ - if (unlikely(!buf_len)) return -EINVAL; - if (unlikely(sock->state == SS_DISCONNECTING)) { - if (skb_queue_len(&sock->sk->sk_receive_queue) == 0) - return -ENOTCONN; - } else if (unlikely(sock->state != SS_CONNECTED)) - return -ENOTCONN; + lock_sock(sk); - /* Look for a message in receive queue; wait if necessary */ - - if (unlikely(down_interruptible(&tsock->sem))) - return -ERESTARTSYS; - -restart: - if (unlikely((skb_queue_len(&sock->sk->sk_receive_queue) == 0) && - (flags & MSG_DONTWAIT))) { - res = -EWOULDBLOCK; - goto exit; - } - - if ((res = wait_event_interruptible( - *sock->sk->sk_sleep, - ((q_len = skb_queue_len(&sock->sk->sk_receive_queue)) || - (sock->state == SS_DISCONNECTING))) )) { + if (unlikely(sock->state == SS_UNCONNECTED)) { + res = -ENOTCONN; goto exit; } - /* Catch attempt to receive on an already terminated connection */ - /* [THIS CHECK MAY OVERLAP WITH AN EARLIER CHECK] */ + target = sock_rcvlowat(sk, flags & MSG_WAITALL, buf_len); + timeo = sock_rcvtimeo(sk, flags & MSG_DONTWAIT); - if (!q_len) { - res = -ENOTCONN; +restart: + /* Look for a message in receive queue; wait if necessary */ + res = tipc_wait_for_rcvmsg(sock, &timeo); + if (res) goto exit; - } - /* Get access to first message in receive queue */ - - buf = skb_peek(&sock->sk->sk_receive_queue); + /* Look at first message in receive queue */ + buf = skb_peek(&sk->sk_receive_queue); msg = buf_msg(buf); sz = msg_data_sz(msg); err = msg_errcode(msg); /* Discard an empty non-errored message & try again */ - if ((!sz) && (!err)) { - advance_queue(tsock); + advance_rx_queue(sk); goto restart; } /* Optionally capture sender's address & ancillary data of first msg */ - if (sz_copied == 0) { set_orig_addr(m, msg); - if ((res = anc_data_recv(m, msg, tsock->p))) + res = anc_data_recv(m, msg, port); + if (res) goto exit; } /* Capture message data (if valid) & compute return value (always) */ - if (!err) { - buf_crs = (unsigned char *)(TIPC_SKB_CB(buf)->handle); - sz = skb_tail_pointer(buf) - buf_crs; + u32 offset = (u32)(unsigned long)(TIPC_SKB_CB(buf)->handle); + sz -= offset; needed = (buf_len - sz_copied); sz_to_copy = (sz <= needed) ? sz : needed; - if (unlikely(copy_to_user(crs, buf_crs, sz_to_copy))) { - res = -EFAULT; + + res = skb_copy_datagram_iovec(buf, msg_hdr_sz(msg) + offset, + m->msg_iov, sz_to_copy); + if (res) goto exit; - } + sz_copied += sz_to_copy; if (sz_to_copy < sz) { if (!(flags & MSG_PEEK)) - TIPC_SKB_CB(buf)->handle = buf_crs + sz_to_copy; + TIPC_SKB_CB(buf)->handle = + (void *)(unsigned long)(offset + sz_to_copy); goto exit; } - - crs += sz_to_copy; } else { if (sz_copied != 0) goto exit; /* can't add error msg to valid data */ @@ -1060,421 +1212,588 @@ restart: } /* Consume received message (optional) */ - if (likely(!(flags & MSG_PEEK))) { - if (unlikely(++tsock->p->conn_unacked >= TIPC_FLOW_CONTROL_WIN)) - tipc_acknowledge(tsock->p->ref, tsock->p->conn_unacked); - advance_queue(tsock); + if (unlikely(++port->conn_unacked >= TIPC_CONNACK_INTV)) + tipc_acknowledge(port->ref, port->conn_unacked); + advance_rx_queue(sk); } /* Loop around if more data is required */ - - if ((sz_copied < buf_len) /* didn't get all requested data */ - && (flags & MSG_WAITALL) /* ... and need to wait for more */ - && (!(flags & MSG_PEEK)) /* ... and aren't just peeking at data */ - && (!err) /* ... and haven't reached a FIN */ - ) + if ((sz_copied < buf_len) && /* didn't get all requested data */ + (!skb_queue_empty(&sk->sk_receive_queue) || + (sz_copied < target)) && /* and more is ready or required */ + (!(flags & MSG_PEEK)) && /* and aren't just peeking at data */ + (!err)) /* and haven't reached a FIN */ goto restart; exit: - up(&tsock->sem); + release_sock(sk); return sz_copied ? sz_copied : res; } /** - * queue_overloaded - test if queue overload condition exists - * @queue_size: current size of queue - * @base: nominal maximum size of queue - * @msg: message to be added to queue - * - * Returns 1 if queue is currently overloaded, 0 otherwise + * tipc_write_space - wake up thread if port congestion is released + * @sk: socket + */ +static void tipc_write_space(struct sock *sk) +{ + struct socket_wq *wq; + + rcu_read_lock(); + wq = rcu_dereference(sk->sk_wq); + if (wq_has_sleeper(wq)) + wake_up_interruptible_sync_poll(&wq->wait, POLLOUT | + POLLWRNORM | POLLWRBAND); + rcu_read_unlock(); +} + +/** + * tipc_data_ready - wake up threads to indicate messages have been received + * @sk: socket + * @len: the length of messages */ +static void tipc_data_ready(struct sock *sk) +{ + struct socket_wq *wq; + + rcu_read_lock(); + wq = rcu_dereference(sk->sk_wq); + if (wq_has_sleeper(wq)) + wake_up_interruptible_sync_poll(&wq->wait, POLLIN | + POLLRDNORM | POLLRDBAND); + rcu_read_unlock(); +} -static int queue_overloaded(u32 queue_size, u32 base, struct tipc_msg *msg) +/** + * filter_connect - Handle all incoming messages for a connection-based socket + * @tsk: TIPC socket + * @msg: message + * + * Returns TIPC error status code and socket error status code + * once it encounters some errors + */ +static u32 filter_connect(struct tipc_sock *tsk, struct sk_buff **buf) { - u32 threshold; - u32 imp = msg_importance(msg); - - if (imp == TIPC_LOW_IMPORTANCE) - threshold = base; - else if (imp == TIPC_MEDIUM_IMPORTANCE) - threshold = base * 2; - else if (imp == TIPC_HIGH_IMPORTANCE) - threshold = base * 100; - else - return 0; + struct sock *sk = &tsk->sk; + struct tipc_port *port = &tsk->port; + struct socket *sock = sk->sk_socket; + struct tipc_msg *msg = buf_msg(*buf); - if (msg_connected(msg)) - threshold *= 4; + u32 retval = TIPC_ERR_NO_PORT; + int res; + + if (msg_mcast(msg)) + return retval; + + switch ((int)sock->state) { + case SS_CONNECTED: + /* Accept only connection-based messages sent by peer */ + if (msg_connected(msg) && tipc_port_peer_msg(port, msg)) { + if (unlikely(msg_errcode(msg))) { + sock->state = SS_DISCONNECTING; + __tipc_port_disconnect(port); + } + retval = TIPC_OK; + } + break; + case SS_CONNECTING: + /* Accept only ACK or NACK message */ + if (unlikely(msg_errcode(msg))) { + sock->state = SS_DISCONNECTING; + sk->sk_err = ECONNREFUSED; + retval = TIPC_OK; + break; + } + + if (unlikely(!msg_connected(msg))) + break; - return (queue_size > threshold); + res = auto_connect(tsk, msg); + if (res) { + sock->state = SS_DISCONNECTING; + sk->sk_err = -res; + retval = TIPC_OK; + break; + } + + /* If an incoming message is an 'ACK-', it should be + * discarded here because it doesn't contain useful + * data. In addition, we should try to wake up + * connect() routine if sleeping. + */ + if (msg_data_sz(msg) == 0) { + kfree_skb(*buf); + *buf = NULL; + if (waitqueue_active(sk_sleep(sk))) + wake_up_interruptible(sk_sleep(sk)); + } + retval = TIPC_OK; + break; + case SS_LISTENING: + case SS_UNCONNECTED: + /* Accept only SYN message */ + if (!msg_connected(msg) && !(msg_errcode(msg))) + retval = TIPC_OK; + break; + case SS_DISCONNECTING: + break; + default: + pr_err("Unknown socket state %u\n", sock->state); + } + return retval; } /** - * async_disconnect - wrapper function used to disconnect port - * @portref: TIPC port reference (passed as pointer-sized value) + * rcvbuf_limit - get proper overload limit of socket receive queue + * @sk: socket + * @buf: message + * + * For all connection oriented messages, irrespective of importance, + * the default overload value (i.e. 67MB) is set as limit. + * + * For all connectionless messages, by default new queue limits are + * as belows: + * + * TIPC_LOW_IMPORTANCE (4 MB) + * TIPC_MEDIUM_IMPORTANCE (8 MB) + * TIPC_HIGH_IMPORTANCE (16 MB) + * TIPC_CRITICAL_IMPORTANCE (32 MB) + * + * Returns overload limit according to corresponding message importance */ - -static void async_disconnect(unsigned long portref) +static unsigned int rcvbuf_limit(struct sock *sk, struct sk_buff *buf) { - tipc_disconnect((u32)portref); + struct tipc_msg *msg = buf_msg(buf); + + if (msg_connected(msg)) + return sysctl_tipc_rmem[2]; + + return sk->sk_rcvbuf >> TIPC_CRITICAL_IMPORTANCE << + msg_importance(msg); } /** - * dispatch - handle arriving message - * @tport: TIPC port that received message + * filter_rcv - validate incoming message + * @sk: socket * @buf: message * - * Called with port locked. Must not take socket lock to avoid deadlock risk. + * Enqueues message on receive queue if acceptable; optionally handles + * disconnect indication for a connected socket. + * + * Called with socket lock already taken; port lock may also be taken. * * Returns TIPC error status code (TIPC_OK if message is not to be rejected) */ - -static u32 dispatch(struct tipc_port *tport, struct sk_buff *buf) +static u32 filter_rcv(struct sock *sk, struct sk_buff *buf) { + struct socket *sock = sk->sk_socket; + struct tipc_sock *tsk = tipc_sk(sk); struct tipc_msg *msg = buf_msg(buf); - struct tipc_sock *tsock = (struct tipc_sock *)tport->usr_handle; - struct socket *sock; - u32 recv_q_len; - - /* Reject message if socket is closing */ - - if (!tsock) - return TIPC_ERR_NO_PORT; + unsigned int limit = rcvbuf_limit(sk, buf); + u32 res = TIPC_OK; /* Reject message if it is wrong sort of message for socket */ + if (msg_type(msg) > TIPC_DIRECT_MSG) + return TIPC_ERR_NO_PORT; - /* - * WOULD IT BE BETTER TO JUST DISCARD THESE MESSAGES INSTEAD? - * "NO PORT" ISN'T REALLY THE RIGHT ERROR CODE, AND THERE MAY - * BE SECURITY IMPLICATIONS INHERENT IN REJECTING INVALID TRAFFIC - */ - sock = tsock->sk.sk_socket; if (sock->state == SS_READY) { - if (msg_connected(msg)) { - msg_dbg(msg, "dispatch filter 1\n"); + if (msg_connected(msg)) return TIPC_ERR_NO_PORT; - } } else { - if (msg_mcast(msg)) { - msg_dbg(msg, "dispatch filter 2\n"); - return TIPC_ERR_NO_PORT; - } - if (sock->state == SS_CONNECTED) { - if (!msg_connected(msg)) { - msg_dbg(msg, "dispatch filter 3\n"); - return TIPC_ERR_NO_PORT; - } - } - else if (sock->state == SS_CONNECTING) { - if (!msg_connected(msg) && (msg_errcode(msg) == 0)) { - msg_dbg(msg, "dispatch filter 4\n"); - return TIPC_ERR_NO_PORT; - } - } - else if (sock->state == SS_LISTENING) { - if (msg_connected(msg) || msg_errcode(msg)) { - msg_dbg(msg, "dispatch filter 5\n"); - return TIPC_ERR_NO_PORT; - } - } - else if (sock->state == SS_DISCONNECTING) { - msg_dbg(msg, "dispatch filter 6\n"); - return TIPC_ERR_NO_PORT; - } - else /* (sock->state == SS_UNCONNECTED) */ { - if (msg_connected(msg) || msg_errcode(msg)) { - msg_dbg(msg, "dispatch filter 7\n"); - return TIPC_ERR_NO_PORT; - } - } + res = filter_connect(tsk, &buf); + if (res != TIPC_OK || buf == NULL) + return res; } /* Reject message if there isn't room to queue it */ + if (sk_rmem_alloc_get(sk) + buf->truesize >= limit) + return TIPC_ERR_OVERLOAD; - if (unlikely((u32)atomic_read(&tipc_queue_size) > - OVERLOAD_LIMIT_BASE)) { - if (queue_overloaded(atomic_read(&tipc_queue_size), - OVERLOAD_LIMIT_BASE, msg)) - return TIPC_ERR_OVERLOAD; - } - recv_q_len = skb_queue_len(&tsock->sk.sk_receive_queue); - if (unlikely(recv_q_len > (OVERLOAD_LIMIT_BASE / 2))) { - if (queue_overloaded(recv_q_len, - OVERLOAD_LIMIT_BASE / 2, msg)) - return TIPC_ERR_OVERLOAD; - } + /* Enqueue message */ + TIPC_SKB_CB(buf)->handle = NULL; + __skb_queue_tail(&sk->sk_receive_queue, buf); + skb_set_owner_r(buf, sk); - /* Initiate connection termination for an incoming 'FIN' */ + sk->sk_data_ready(sk); + return TIPC_OK; +} - if (unlikely(msg_errcode(msg) && (sock->state == SS_CONNECTED))) { - sock->state = SS_DISCONNECTING; - /* Note: Use signal since port lock is already taken! */ - tipc_k_signal((Handler)async_disconnect, tport->ref); - } +/** + * tipc_backlog_rcv - handle incoming message from backlog queue + * @sk: socket + * @buf: message + * + * Caller must hold socket lock, but not port lock. + * + * Returns 0 + */ +static int tipc_backlog_rcv(struct sock *sk, struct sk_buff *buf) +{ + u32 res; + struct tipc_sock *tsk = tipc_sk(sk); + uint truesize = buf->truesize; - /* Enqueue message (finally!) */ + res = filter_rcv(sk, buf); + if (unlikely(res)) + tipc_reject_msg(buf, res); - msg_dbg(msg,"<DISP<: "); - TIPC_SKB_CB(buf)->handle = msg_data(msg); - atomic_inc(&tipc_queue_size); - skb_queue_tail(&sock->sk->sk_receive_queue, buf); + if (atomic_read(&tsk->dupl_rcvcnt) < TIPC_CONN_OVERLOAD_LIMIT) + atomic_add(truesize, &tsk->dupl_rcvcnt); - if (waitqueue_active(sock->sk->sk_sleep)) - wake_up_interruptible(sock->sk->sk_sleep); - return TIPC_OK; + return 0; } /** - * wakeupdispatch - wake up port after congestion - * @tport: port to wakeup - * - * Called with port lock on. + * tipc_sk_rcv - handle incoming message + * @buf: buffer containing arriving message + * Consumes buffer + * Returns 0 if success, or errno: -EHOSTUNREACH */ +int tipc_sk_rcv(struct sk_buff *buf) +{ + struct tipc_sock *tsk; + struct tipc_port *port; + struct sock *sk; + u32 dport = msg_destport(buf_msg(buf)); + int err = TIPC_OK; + uint limit; + + /* Forward unresolved named message */ + if (unlikely(!dport)) { + tipc_net_route_msg(buf); + return 0; + } + + /* Validate destination */ + port = tipc_port_lock(dport); + if (unlikely(!port)) { + err = TIPC_ERR_NO_PORT; + goto exit; + } + + tsk = tipc_port_to_sock(port); + sk = &tsk->sk; + + /* Queue message */ + bh_lock_sock(sk); + + if (!sock_owned_by_user(sk)) { + err = filter_rcv(sk, buf); + } else { + if (sk->sk_backlog.len == 0) + atomic_set(&tsk->dupl_rcvcnt, 0); + limit = rcvbuf_limit(sk, buf) + atomic_read(&tsk->dupl_rcvcnt); + if (sk_add_backlog(sk, buf, limit)) + err = TIPC_ERR_OVERLOAD; + } + + bh_unlock_sock(sk); + tipc_port_unlock(port); + + if (likely(!err)) + return 0; +exit: + tipc_reject_msg(buf, err); + return -EHOSTUNREACH; +} -static void wakeupdispatch(struct tipc_port *tport) +static int tipc_wait_for_connect(struct socket *sock, long *timeo_p) { - struct tipc_sock *tsock = (struct tipc_sock *)tport->usr_handle; + struct sock *sk = sock->sk; + DEFINE_WAIT(wait); + int done; - if (waitqueue_active(tsock->sk.sk_sleep)) - wake_up_interruptible(tsock->sk.sk_sleep); + do { + int err = sock_error(sk); + if (err) + return err; + if (!*timeo_p) + return -ETIMEDOUT; + if (signal_pending(current)) + return sock_intr_errno(*timeo_p); + + prepare_to_wait(sk_sleep(sk), &wait, TASK_INTERRUPTIBLE); + done = sk_wait_event(sk, timeo_p, sock->state != SS_CONNECTING); + finish_wait(sk_sleep(sk), &wait); + } while (!done); + return 0; } /** - * connect - establish a connection to another TIPC port + * tipc_connect - establish a connection to another TIPC port * @sock: socket structure * @dest: socket address for destination port * @destlen: size of socket address data structure - * @flags: (unused) + * @flags: file-related flags associated with socket * * Returns 0 on success, errno otherwise */ - -static int connect(struct socket *sock, struct sockaddr *dest, int destlen, - int flags) +static int tipc_connect(struct socket *sock, struct sockaddr *dest, + int destlen, int flags) { - struct tipc_sock *tsock = tipc_sk(sock->sk); - struct sockaddr_tipc *dst = (struct sockaddr_tipc *)dest; - struct msghdr m = {NULL,}; - struct sk_buff *buf; - struct tipc_msg *msg; - int res; - - /* For now, TIPC does not allow use of connect() with DGRAM or RDM types */ - - if (sock->state == SS_READY) - return -EOPNOTSUPP; - - /* Issue Posix-compliant error code if socket is in the wrong state */ - - if (sock->state == SS_LISTENING) - return -EOPNOTSUPP; - if (sock->state == SS_CONNECTING) - return -EALREADY; - if (sock->state != SS_UNCONNECTED) - return -EISCONN; - - /* - * Reject connection attempt using multicast address - * - * Note: send_msg() validates the rest of the address fields, - * so there's no need to do it here - */ - - if (dst->addrtype == TIPC_ADDR_MCAST) - return -EINVAL; - - /* Send a 'SYN-' to destination */ - - m.msg_name = dest; - m.msg_namelen = destlen; - if ((res = send_msg(NULL, sock, &m, 0)) < 0) { - sock->state = SS_DISCONNECTING; - return res; - } - - if (down_interruptible(&tsock->sem)) - return -ERESTARTSYS; - - /* Wait for destination's 'ACK' response */ - - res = wait_event_interruptible_timeout(*sock->sk->sk_sleep, - skb_queue_len(&sock->sk->sk_receive_queue), - sock->sk->sk_rcvtimeo); - buf = skb_peek(&sock->sk->sk_receive_queue); - if (res > 0) { - msg = buf_msg(buf); - res = auto_connect(sock, tsock, msg); - if (!res) { - if (!msg_data_sz(msg)) - advance_queue(tsock); - } - } else { - if (res == 0) { - res = -ETIMEDOUT; - } else - { /* leave "res" unchanged */ } - sock->state = SS_DISCONNECTING; - } - - up(&tsock->sem); - return res; + struct sock *sk = sock->sk; + struct sockaddr_tipc *dst = (struct sockaddr_tipc *)dest; + struct msghdr m = {NULL,}; + long timeout = (flags & O_NONBLOCK) ? 0 : tipc_sk(sk)->conn_timeout; + socket_state previous; + int res; + + lock_sock(sk); + + /* For now, TIPC does not allow use of connect() with DGRAM/RDM types */ + if (sock->state == SS_READY) { + res = -EOPNOTSUPP; + goto exit; + } + + /* + * Reject connection attempt using multicast address + * + * Note: send_msg() validates the rest of the address fields, + * so there's no need to do it here + */ + if (dst->addrtype == TIPC_ADDR_MCAST) { + res = -EINVAL; + goto exit; + } + + previous = sock->state; + switch (sock->state) { + case SS_UNCONNECTED: + /* Send a 'SYN-' to destination */ + m.msg_name = dest; + m.msg_namelen = destlen; + + /* If connect is in non-blocking case, set MSG_DONTWAIT to + * indicate send_msg() is never blocked. + */ + if (!timeout) + m.msg_flags = MSG_DONTWAIT; + + res = tipc_sendmsg(NULL, sock, &m, 0); + if ((res < 0) && (res != -EWOULDBLOCK)) + goto exit; + + /* Just entered SS_CONNECTING state; the only + * difference is that return value in non-blocking + * case is EINPROGRESS, rather than EALREADY. + */ + res = -EINPROGRESS; + case SS_CONNECTING: + if (previous == SS_CONNECTING) + res = -EALREADY; + if (!timeout) + goto exit; + timeout = msecs_to_jiffies(timeout); + /* Wait until an 'ACK' or 'RST' arrives, or a timeout occurs */ + res = tipc_wait_for_connect(sock, &timeout); + break; + case SS_CONNECTED: + res = -EISCONN; + break; + default: + res = -EINVAL; + break; + } +exit: + release_sock(sk); + return res; } /** - * listen - allow socket to listen for incoming connections + * tipc_listen - allow socket to listen for incoming connections * @sock: socket structure * @len: (unused) * * Returns 0 on success, errno otherwise */ - -static int listen(struct socket *sock, int len) +static int tipc_listen(struct socket *sock, int len) { - /* REQUIRES SOCKET LOCKING OF SOME SORT? */ + struct sock *sk = sock->sk; + int res; + + lock_sock(sk); - if (sock->state == SS_READY) - return -EOPNOTSUPP; if (sock->state != SS_UNCONNECTED) - return -EINVAL; - sock->state = SS_LISTENING; - return 0; + res = -EINVAL; + else { + sock->state = SS_LISTENING; + res = 0; + } + + release_sock(sk); + return res; +} + +static int tipc_wait_for_accept(struct socket *sock, long timeo) +{ + struct sock *sk = sock->sk; + DEFINE_WAIT(wait); + int err; + + /* True wake-one mechanism for incoming connections: only + * one process gets woken up, not the 'whole herd'. + * Since we do not 'race & poll' for established sockets + * anymore, the common case will execute the loop only once. + */ + for (;;) { + prepare_to_wait_exclusive(sk_sleep(sk), &wait, + TASK_INTERRUPTIBLE); + if (timeo && skb_queue_empty(&sk->sk_receive_queue)) { + release_sock(sk); + timeo = schedule_timeout(timeo); + lock_sock(sk); + } + err = 0; + if (!skb_queue_empty(&sk->sk_receive_queue)) + break; + err = -EINVAL; + if (sock->state != SS_LISTENING) + break; + err = sock_intr_errno(timeo); + if (signal_pending(current)) + break; + err = -EAGAIN; + if (!timeo) + break; + } + finish_wait(sk_sleep(sk), &wait); + return err; } /** - * accept - wait for connection request + * tipc_accept - wait for connection request * @sock: listening socket * @newsock: new socket that is to be connected * @flags: file-related flags associated with socket * * Returns 0 on success, errno otherwise */ - -static int accept(struct socket *sock, struct socket *newsock, int flags) +static int tipc_accept(struct socket *sock, struct socket *new_sock, int flags) { - struct tipc_sock *tsock = tipc_sk(sock->sk); + struct sock *new_sk, *sk = sock->sk; struct sk_buff *buf; - int res = -EFAULT; + struct tipc_port *new_port; + struct tipc_msg *msg; + struct tipc_portid peer; + u32 new_ref; + long timeo; + int res; - if (sock->state == SS_READY) - return -EOPNOTSUPP; - if (sock->state != SS_LISTENING) - return -EINVAL; + lock_sock(sk); - if (unlikely((skb_queue_len(&sock->sk->sk_receive_queue) == 0) && - (flags & O_NONBLOCK))) - return -EWOULDBLOCK; + if (sock->state != SS_LISTENING) { + res = -EINVAL; + goto exit; + } + timeo = sock_rcvtimeo(sk, flags & O_NONBLOCK); + res = tipc_wait_for_accept(sock, timeo); + if (res) + goto exit; - if (down_interruptible(&tsock->sem)) - return -ERESTARTSYS; + buf = skb_peek(&sk->sk_receive_queue); - if (wait_event_interruptible(*sock->sk->sk_sleep, - skb_queue_len(&sock->sk->sk_receive_queue))) { - res = -ERESTARTSYS; + res = tipc_sk_create(sock_net(sock->sk), new_sock, 0, 1); + if (res) goto exit; - } - buf = skb_peek(&sock->sk->sk_receive_queue); - - res = tipc_create(sock->sk->sk_net, newsock, 0); - if (!res) { - struct tipc_sock *new_tsock = tipc_sk(newsock->sk); - struct tipc_portid id; - struct tipc_msg *msg = buf_msg(buf); - u32 new_ref = new_tsock->p->ref; - - id.ref = msg_origport(msg); - id.node = msg_orignode(msg); - tipc_connect2port(new_ref, &id); - newsock->state = SS_CONNECTED; - - tipc_set_portimportance(new_ref, msg_importance(msg)); - if (msg_named(msg)) { - new_tsock->p->conn_type = msg_nametype(msg); - new_tsock->p->conn_instance = msg_nameinst(msg); - } - /* - * Respond to 'SYN-' by discarding it & returning 'ACK'-. - * Respond to 'SYN+' by queuing it on new socket. - */ + new_sk = new_sock->sk; + new_port = &tipc_sk(new_sk)->port; + new_ref = new_port->ref; + msg = buf_msg(buf); - msg_dbg(msg,"<ACC<: "); - if (!msg_data_sz(msg)) { - struct msghdr m = {NULL,}; + /* we lock on new_sk; but lockdep sees the lock on sk */ + lock_sock_nested(new_sk, SINGLE_DEPTH_NESTING); - send_packet(NULL, newsock, &m, 0); - advance_queue(tsock); - } else { - sock_lock(tsock); - skb_dequeue(&sock->sk->sk_receive_queue); - sock_unlock(tsock); - skb_queue_head(&newsock->sk->sk_receive_queue, buf); - } + /* + * Reject any stray messages received by new socket + * before the socket lock was taken (very, very unlikely) + */ + reject_rx_queue(new_sk); + + /* Connect new socket to it's peer */ + peer.ref = msg_origport(msg); + peer.node = msg_orignode(msg); + tipc_port_connect(new_ref, &peer); + new_sock->state = SS_CONNECTED; + + tipc_port_set_importance(new_port, msg_importance(msg)); + if (msg_named(msg)) { + new_port->conn_type = msg_nametype(msg); + new_port->conn_instance = msg_nameinst(msg); } + + /* + * Respond to 'SYN-' by discarding it & returning 'ACK'-. + * Respond to 'SYN+' by queuing it on new socket. + */ + if (!msg_data_sz(msg)) { + struct msghdr m = {NULL,}; + + advance_rx_queue(sk); + tipc_send_packet(NULL, new_sock, &m, 0); + } else { + __skb_dequeue(&sk->sk_receive_queue); + __skb_queue_head(&new_sk->sk_receive_queue, buf); + skb_set_owner_r(buf, new_sk); + } + release_sock(new_sk); exit: - up(&tsock->sem); + release_sock(sk); return res; } /** - * shutdown - shutdown socket connection + * tipc_shutdown - shutdown socket connection * @sock: socket structure - * @how: direction to close (unused; always treated as read + write) + * @how: direction to close (must be SHUT_RDWR) * * Terminates connection (if necessary), then purges socket's receive queue. * * Returns 0 on success, errno otherwise */ - -static int shutdown(struct socket *sock, int how) +static int tipc_shutdown(struct socket *sock, int how) { - struct tipc_sock* tsock = tipc_sk(sock->sk); + struct sock *sk = sock->sk; + struct tipc_sock *tsk = tipc_sk(sk); + struct tipc_port *port = &tsk->port; struct sk_buff *buf; int res; - /* Could return -EINVAL for an invalid "how", but why bother? */ - - if (down_interruptible(&tsock->sem)) - return -ERESTARTSYS; + if (how != SHUT_RDWR) + return -EINVAL; - sock_lock(tsock); + lock_sock(sk); switch (sock->state) { + case SS_CONNECTING: case SS_CONNECTED: - /* Send 'FIN+' or 'FIN-' message to peer */ - - sock_unlock(tsock); restart: - if ((buf = skb_dequeue(&sock->sk->sk_receive_queue))) { - atomic_dec(&tipc_queue_size); - if (TIPC_SKB_CB(buf)->handle != msg_data(buf_msg(buf))) { - buf_discard(buf); + /* Disconnect and send a 'FIN+' or 'FIN-' message to peer */ + buf = __skb_dequeue(&sk->sk_receive_queue); + if (buf) { + if (TIPC_SKB_CB(buf)->handle != NULL) { + kfree_skb(buf); goto restart; } + tipc_port_disconnect(port->ref); tipc_reject_msg(buf, TIPC_CONN_SHUTDOWN); + } else { + tipc_port_shutdown(port->ref); } - else { - tipc_shutdown(tsock->p->ref); - } - sock_lock(tsock); + + sock->state = SS_DISCONNECTING; /* fall through */ case SS_DISCONNECTING: /* Discard any unreceived messages */ + __skb_queue_purge(&sk->sk_receive_queue); - while ((buf = skb_dequeue(&sock->sk->sk_receive_queue))) { - atomic_dec(&tipc_queue_size); - buf_discard(buf); - } - tsock->p->conn_unacked = 0; - - /* fall through */ - - case SS_CONNECTING: - sock->state = SS_DISCONNECTING; + /* Wake up anyone sleeping in poll */ + sk->sk_state_change(sk); res = 0; break; @@ -1482,14 +1801,12 @@ restart: res = -ENOTCONN; } - sock_unlock(tsock); - - up(&tsock->sem); + release_sock(sk); return res; } /** - * setsockopt - set socket option + * tipc_setsockopt - set socket option * @sock: socket structure * @lvl: option level * @opt: option identifier @@ -1501,11 +1818,12 @@ restart: * * Returns 0 on success, errno otherwise */ - -static int setsockopt(struct socket *sock, - int lvl, int opt, char __user *ov, int ol) +static int tipc_setsockopt(struct socket *sock, int lvl, int opt, + char __user *ov, unsigned int ol) { - struct tipc_sock *tsock = tipc_sk(sock->sk); + struct sock *sk = sock->sk; + struct tipc_sock *tsk = tipc_sk(sk); + struct tipc_port *port = &tsk->port; u32 value; int res; @@ -1515,38 +1833,40 @@ static int setsockopt(struct socket *sock, return -ENOPROTOOPT; if (ol < sizeof(value)) return -EINVAL; - if ((res = get_user(value, (u32 __user *)ov))) + res = get_user(value, (u32 __user *)ov); + if (res) return res; - if (down_interruptible(&tsock->sem)) - return -ERESTARTSYS; + lock_sock(sk); switch (opt) { case TIPC_IMPORTANCE: - res = tipc_set_portimportance(tsock->p->ref, value); + tipc_port_set_importance(port, value); break; case TIPC_SRC_DROPPABLE: if (sock->type != SOCK_STREAM) - res = tipc_set_portunreliable(tsock->p->ref, value); + tipc_port_set_unreliable(port, value); else res = -ENOPROTOOPT; break; case TIPC_DEST_DROPPABLE: - res = tipc_set_portunreturnable(tsock->p->ref, value); + tipc_port_set_unreturnable(port, value); break; case TIPC_CONN_TIMEOUT: - sock->sk->sk_rcvtimeo = (value * HZ / 1000); + tipc_sk(sk)->conn_timeout = value; + /* no need to set "res", since already 0 at this point */ break; default: res = -EINVAL; } - up(&tsock->sem); + release_sock(sk); + return res; } /** - * getsockopt - get socket option + * tipc_getsockopt - get socket option * @sock: socket structure * @lvl: option level * @opt: option identifier @@ -1558,11 +1878,12 @@ static int setsockopt(struct socket *sock, * * Returns 0 on success, errno otherwise */ - -static int getsockopt(struct socket *sock, - int lvl, int opt, char __user *ov, int __user *ol) +static int tipc_getsockopt(struct socket *sock, int lvl, int opt, + char __user *ov, int __user *ol) { - struct tipc_sock *tsock = tipc_sk(sock->sk); + struct sock *sk = sock->sk; + struct tipc_sock *tsk = tipc_sk(sk); + struct tipc_port *port = &tsk->port; int len; u32 value; int res; @@ -1571,123 +1892,154 @@ static int getsockopt(struct socket *sock, return put_user(0, ol); if (lvl != SOL_TIPC) return -ENOPROTOOPT; - if ((res = get_user(len, ol))) + res = get_user(len, ol); + if (res) return res; - if (down_interruptible(&tsock->sem)) - return -ERESTARTSYS; + lock_sock(sk); switch (opt) { case TIPC_IMPORTANCE: - res = tipc_portimportance(tsock->p->ref, &value); + value = tipc_port_importance(port); break; case TIPC_SRC_DROPPABLE: - res = tipc_portunreliable(tsock->p->ref, &value); + value = tipc_port_unreliable(port); break; case TIPC_DEST_DROPPABLE: - res = tipc_portunreturnable(tsock->p->ref, &value); + value = tipc_port_unreturnable(port); break; case TIPC_CONN_TIMEOUT: - value = (sock->sk->sk_rcvtimeo * 1000) / HZ; + value = tipc_sk(sk)->conn_timeout; + /* no need to set "res", since already 0 at this point */ + break; + case TIPC_NODE_RECVQ_DEPTH: + value = 0; /* was tipc_queue_size, now obsolete */ + break; + case TIPC_SOCK_RECVQ_DEPTH: + value = skb_queue_len(&sk->sk_receive_queue); break; default: res = -EINVAL; } - if (res) { - /* "get" failed */ - } - else if (len < sizeof(value)) { - res = -EINVAL; - } - else if ((res = copy_to_user(ov, &value, sizeof(value)))) { - /* couldn't return value */ - } - else { - res = put_user(sizeof(value), ol); - } + release_sock(sk); - up(&tsock->sem); - return res; + if (res) + return res; /* "get" failed */ + + if (len < sizeof(value)) + return -EINVAL; + + if (copy_to_user(ov, &value, sizeof(value))) + return -EFAULT; + + return put_user(sizeof(value), ol); } -/** - * Protocol switches for the various types of TIPC sockets - */ +int tipc_ioctl(struct socket *sk, unsigned int cmd, unsigned long arg) +{ + struct tipc_sioc_ln_req lnr; + void __user *argp = (void __user *)arg; + + switch (cmd) { + case SIOCGETLINKNAME: + if (copy_from_user(&lnr, argp, sizeof(lnr))) + return -EFAULT; + if (!tipc_node_get_linkname(lnr.bearer_id, lnr.peer, + lnr.linkname, TIPC_MAX_LINK_NAME)) { + if (copy_to_user(argp, &lnr, sizeof(lnr))) + return -EFAULT; + return 0; + } + return -EADDRNOTAVAIL; + break; + default: + return -ENOIOCTLCMD; + } +} + +/* Protocol switches for the various types of TIPC sockets */ static const struct proto_ops msg_ops = { - .owner = THIS_MODULE, + .owner = THIS_MODULE, .family = AF_TIPC, - .release = release, - .bind = bind, - .connect = connect, + .release = tipc_release, + .bind = tipc_bind, + .connect = tipc_connect, .socketpair = sock_no_socketpair, - .accept = accept, - .getname = get_name, - .poll = poll, - .ioctl = sock_no_ioctl, - .listen = listen, - .shutdown = shutdown, - .setsockopt = setsockopt, - .getsockopt = getsockopt, - .sendmsg = send_msg, - .recvmsg = recv_msg, + .accept = sock_no_accept, + .getname = tipc_getname, + .poll = tipc_poll, + .ioctl = tipc_ioctl, + .listen = sock_no_listen, + .shutdown = tipc_shutdown, + .setsockopt = tipc_setsockopt, + .getsockopt = tipc_getsockopt, + .sendmsg = tipc_sendmsg, + .recvmsg = tipc_recvmsg, .mmap = sock_no_mmap, .sendpage = sock_no_sendpage }; static const struct proto_ops packet_ops = { - .owner = THIS_MODULE, + .owner = THIS_MODULE, .family = AF_TIPC, - .release = release, - .bind = bind, - .connect = connect, + .release = tipc_release, + .bind = tipc_bind, + .connect = tipc_connect, .socketpair = sock_no_socketpair, - .accept = accept, - .getname = get_name, - .poll = poll, - .ioctl = sock_no_ioctl, - .listen = listen, - .shutdown = shutdown, - .setsockopt = setsockopt, - .getsockopt = getsockopt, - .sendmsg = send_packet, - .recvmsg = recv_msg, + .accept = tipc_accept, + .getname = tipc_getname, + .poll = tipc_poll, + .ioctl = tipc_ioctl, + .listen = tipc_listen, + .shutdown = tipc_shutdown, + .setsockopt = tipc_setsockopt, + .getsockopt = tipc_getsockopt, + .sendmsg = tipc_send_packet, + .recvmsg = tipc_recvmsg, .mmap = sock_no_mmap, .sendpage = sock_no_sendpage }; static const struct proto_ops stream_ops = { - .owner = THIS_MODULE, + .owner = THIS_MODULE, .family = AF_TIPC, - .release = release, - .bind = bind, - .connect = connect, + .release = tipc_release, + .bind = tipc_bind, + .connect = tipc_connect, .socketpair = sock_no_socketpair, - .accept = accept, - .getname = get_name, - .poll = poll, - .ioctl = sock_no_ioctl, - .listen = listen, - .shutdown = shutdown, - .setsockopt = setsockopt, - .getsockopt = getsockopt, - .sendmsg = send_stream, - .recvmsg = recv_stream, + .accept = tipc_accept, + .getname = tipc_getname, + .poll = tipc_poll, + .ioctl = tipc_ioctl, + .listen = tipc_listen, + .shutdown = tipc_shutdown, + .setsockopt = tipc_setsockopt, + .getsockopt = tipc_getsockopt, + .sendmsg = tipc_send_stream, + .recvmsg = tipc_recv_stream, .mmap = sock_no_mmap, .sendpage = sock_no_sendpage }; static const struct net_proto_family tipc_family_ops = { - .owner = THIS_MODULE, + .owner = THIS_MODULE, .family = AF_TIPC, - .create = tipc_create + .create = tipc_sk_create }; static struct proto tipc_proto = { .name = "TIPC", .owner = THIS_MODULE, - .obj_size = sizeof(struct tipc_sock) + .obj_size = sizeof(struct tipc_sock), + .sysctl_rmem = sysctl_tipc_rmem +}; + +static struct proto tipc_proto_kern = { + .name = "TIPC", + .obj_size = sizeof(struct tipc_sock), + .sysctl_rmem = sysctl_tipc_rmem }; /** @@ -1701,18 +2053,16 @@ int tipc_socket_init(void) res = proto_register(&tipc_proto, 1); if (res) { - err("Failed to register TIPC protocol type\n"); + pr_err("Failed to register TIPC protocol type\n"); goto out; } res = sock_register(&tipc_family_ops); if (res) { - err("Failed to register TIPC socket type\n"); + pr_err("Failed to register TIPC socket type\n"); proto_unregister(&tipc_proto); goto out; } - - sockets_enabled = 1; out: return res; } @@ -1722,11 +2072,6 @@ int tipc_socket_init(void) */ void tipc_socket_stop(void) { - if (!sockets_enabled) - return; - - sockets_enabled = 0; sock_unregister(tipc_family_ops.family); proto_unregister(&tipc_proto); } - |